Prediction of pancreatitis following endoscopic retrograde cholangiopancreatography

OBJECTIVE The aim of this study was to determine if the difference in serum amylase levels prior to, and two hours following, an endoscopic retrograde cholangiopancreatography (ERCP), or the ratio of the two-hour post-ERCP amylase level to the pre-ERCP amylase level was a better predictor of post-ERCP pancreatitis (PEP). METHODS This was a retrospective, single-center study of consecutive patients, who underwent ERCP between April 2015 and August 2018. Serum amylase was measured before and two hours following ERCP. We compared the difference and the ratio of the two levels in predicting PEP using a receiver operating characteristic (ROC) curve analysis. RESULTS A total of 1029 patients underwent ERCP, with PEP occurring in 118 (11.5%). Multivariate analysis revealed that an elevated two-hour post-ERCP serum amylase level was a significant predictor of PEP. ROC analysis of the difference and the ratio of the two levels found good performance for both parameters, with an area under the curve (AUC) of 0.861 (95% confidence interval [CI], 0.823-0.900) and 0.847 (95% CI, 0.809-0.886), respectively. The difference between the values was a significantly more effective predictor of PEP, based on the AUC analysis (P = 0.011). CONCLUSION The difference between pre and two-hour post-ERCP amylase levels is a better predictor of PEP than the ratio of the two.